The use of nurses' and carers' observations in the identification of poststroke depression

Aim. This paper is a report of a study to determine the accuracy and utility of an observational screening tool (Signs of Depression Scale), when rated by nurses and carers for detecting depression in patients who have recently had a stroke. Background. Depression following a stroke is common and adversely affects recovery and rehabilitation. Identifying depression when patients have communication and cognitive difficulties is especially problematic. Screening tools which rely solely on observation may be beneficial in this group. Method. A cross-sectional study of people admitted with an acute stroke compared a clinical diagnosis of depression by a psychiatrist (the gold standard) with the Signs of Depression Scale completed by nurses and carers. The agreement between nurses' and carers' ratings was also explored. Data were collected over 10 months (December 2004-October 2005). Findings. Seventy-one patients were included in the study, median age 70 [inter-quartile ranges (IQR) 59-76], including 40 (56.3%) males. The psychiatrist classified 25/71 (35.2%) patients as depressed. Using the recommended cut-point of 2 or more on the Signs of Depression Scale, the nurse and carer respectively rated 27/71 (38.0%) and 18/30 (60.0%) patients as potentially depressed. The proportion of patients correctly identified as depressed by the test (sensitivity) when rated by nurses was 64%, and the proportion of patients not depressed who were correctly identified by the test (specificity) was 61%, whereas carers achieved sensitivity 90% and specificity 35%. The optimal cut-point for carers was higher at 4 or more. Inter-rater agreement on the Signs of Depression Scale between nurses and carers was fair (ICC = 0.43, 95% CI: 0.09-0.68). Conclusion. The Signs of Depression Scale is easily completed by clinical staff, although we found the sensitivity when completed by nurses to be low. Information from carers shows potential to improve screening and it is important for nurses to value the knowledge and skills of carers in detecting depression following a stroke. Further refinement of the Signs of Depression Scale, with accompanying research, is required. © 2007 The Authors.